Evaluation of a Model to Target High-risk Psychiatric Inpatients for an Intensive Postdischarge Suicide Prevention Intervention.

Importance: The months after psychiatric hospital discharge are a time of high risk for suicide. Intensive postdischarge case management, although potentially effective in suicide prevention, is likely to be cost-effective only if targeted at high-risk patients. A previously developed machine learning (ML) model showed that postdischarge suicides can be predicted from electronic health records and geospatial data, but it is unknown if prediction could be improved by adding additional information. Objective: To determine whether model prediction could be improved by adding information extracted from clinical notes and public records. Design, Setting, and Participants: Models were trained to predict suicides in the 12 months after Veterans Health Administration (VHA) short-term (less than 365 days) psychiatric hospitalizations between the beginning of 2010 and September 1, 2012 (299 050 hospitalizations, with 916 hospitalizations followed within 12 months by suicides) and tested in the hospitalizations from September 2, 2012, to December 31, 2013 (149 738 hospitalizations, with 393 hospitalizations followed within 12 months by suicides). Validation focused on net benefit across a range of plausible decision thresholds. Predictor importance was assessed with Shapley additive explanations (SHAP) values. Data were analyzed from January to August 2022. Main Outcomes and Measures: Suicides were defined by the National Death Index. Base model predictors included VHA electronic health records and patient residential data. The expanded predictors came from natural language processing (NLP) of clinical notes and a social determinants of health (SDOH) public records database. Results: The model included 448 788 unique hospitalizations. Net benefit over risk horizons between 3 and 12 months was generally highest for the model that included both NLP and SDOH predictors (area under the receiver operating characteristic curve range, 0.747-0.780; area under the precision recall curve relative to the suicide rate range, 3.87-5.75). NLP and SDOH predictors also had the highest predictor class-level SHAP values (proportional SHAP = 64.0% and 49.3%, respectively), although the single highest positive variable-level SHAP value was for a count of medications classified by the US Food and Drug Administration as increasing suicide risk prescribed the year before hospitalization (proportional SHAP = 15.0%). Conclusions and Relevance: In this study, clinical notes and public records were found to improve ML model prediction of suicide after psychiatric hospitalization. The model had positive net benefit over 3-month to 12-month risk horizons for plausible decision thresholds. Although caution is needed in inferring causality based on predictor importance, several key predictors have potential intervention implications that should be investigated in future studies.

Under-recognition of medically unexplained symptom conditions among US Veterans with Gulf War Illness.

OBJECTIVE: Conditions defined by persistent "medically unexplained" physical symptoms and syndromes (MUS) are common and disabling. Veterans from the Gulf War (deployed 1990-1991) have notably high prevalence and disability from MUS conditions. Individuals with MUS report that providers do not recognize their MUS conditions. Our goal was to determine if Veterans with MUS receive an ICD-10 diagnosis for a MUS condition or receive disability benefits available to them for these conditions. METHODS: A chart review was conducted with US Veterans who met case criteria for Gulf War Illness, a complex MUS condition (N = 204, M = 53 years-old, SD = 7). Three coders independently reviewed Veteran's medical records for MUS condition diagnosis or service-connection along with comorbid mental and physical health conditions. Service-connection refers to US Veterans Affairs disability benefits eligibility for conditions or injuries experienced during or exacerbated by military service. RESULTS: Twenty-nine percent had a diagnosis of a MUS condition in their medical record, the most common were irritable colon/irritable bowel syndrome (16%) and fibromyalgia (11%). Slightly more Veterans were service-connected for a MUS condition (38%) as compared to diagnosed. There were high rates of diagnoses and service-connection for mental health (diagnoses 76% and service-connection 74%), musculoskeletal (diagnoses 86%, service-connection 79%), and illness-related conditions (diagnoses 98%, service-connection 49%). CONCLUSION: Given that all participants were Gulf War Veterans who met criteria for a MUS condition, our results suggest that MUS conditions in Gulf War Veterans are under-recognized with regard to clinical diagnosis and service-connected disability. Veterans were more likely to be diagnosed and service-connected for musculoskeletal-related and mental health conditions than MUS conditions. Providers may need education and training to facilitate diagnosis of and service-connection for MUS conditions. We believe that greater acknowledgement and validation of MUS conditions would increase patient engagement with healthcare as well as provider and patient satisfaction with care.

Sleep Disturbances and Atopic Dermatitis: Relationships, Methods for Assessment, and Therapies.

Atopic dermatitis is one of the most common chronic inflammatory skin conditions and is associated with sleep disturbances in 47% to 80% of children and 33% to 90% of adults. Herein, we review the literature on sleep disturbances experienced by patients with atopic dermatitis, as well as the mechanisms that may underlie this. We present subjective and objective methods for measuring sleep quantity and quality and discuss strategies for management. Unfortunately, the literature on this topic remains sparse, with most studies evaluating sleep as a secondary outcome using subjective measures. The development of portable, at-home methods for more objective measures offers new opportunities to better evaluate sleep disturbances in atopic dermatitis research studies and in clinical practice.

Using Administrative Data to Predict Suicide After Psychiatric Hospitalization in the Veterans Health Administration System.

There is a very high suicide rate in the year after psychiatric hospital discharge. Intensive postdischarge case management programs can address this problem but are not cost-effective for all patients. This issue can be addressed by developing a risk model to predict which inpatients might need such a program. We developed such a model for the 391,018 short-term psychiatric hospital admissions of US veterans in Veterans Health Administration (VHA) hospitals 2010-2013. Records were linked with the National Death Index to determine suicide within 12 months of hospital discharge (n=771). The Super Learner ensemble machine learning method was used to predict these suicides for time horizon between 1 week and 12 months after discharge in a 70% training sample. Accuracy was validated in the remaining 30% holdout sample. Predictors included VHA administrative variables and small area geocode data linked to patient home addresses. The models had AUC=.79-.82 for time horizons between 1 week and 6 months and AUC=.74 for 12 months. An analysis of operating characteristics showed that 22.4%-32.2% of patients who died by suicide would have been reached if intensive case management was provided to the 5% of patients with highest predicted suicide risk. Positive predictive value (PPV) at this higher threshold ranged from 1.2% over 12 months to 3.8% per case manager year over 1 week. Focusing on the low end of the risk spectrum, the 40% of patients classified as having lowest risk account for 0%-9.7% of suicides across time horizons. Variable importance analysis shows that 51.1% of model performance is due to psychopathological risk factors accounted, 26.2% to social determinants of health, 14.8% to prior history of suicidal behaviors, and 6.6% to physical disorders. The paper closes with a discussion of next steps in refining the model and prospects for developing a parallel precision treatment model.

Considerations in the Use of Interactive Voice Recording for the Temporal Assessment of Suicidal Ideation and Alcohol Use.

BACKGROUND: A greater understanding of the temporal variation of suicidal ideation and suicidal behavior is needed to inform more effective prevention efforts. Interactive voice recording (IVR) allows for the study of temporal relationships that cannot be captured with most traditional methodologies. AIMS: To examine the feasibility of implementing IVR for the assessment of suicidal ideation. METHOD: Participants (n = 4) receiving a brief intervention based on dialectical behavior therapy were asked to respond to three phone-based surveys each day over 6 weeks that assessed suicidal ideation and alcohol consumption. RESULTS: Participants completed 77.7% of daily assessments, reported that calls were not burdensome, and indicated that calls were sometimes helpful in interrupting suicidal ideation. CONCLUSION: The preliminary data reported here provide optimism for the use of IVR and other forms of ecological momentary assessment in the exploration of the antecedents of suicidal behavior.

Elevated sleep disturbance among blacks in an urban family medicine practice.

PURPOSE: Blacks experience a number of health disparities. Sleep disturbances contribute to poor health. This preliminary study explores whether a disparity in sleep disturbances exists among blacks compared with whites and others. METHODS: A cross-sectional study was conducted in a sample (n = 92) of urban primary care patients (52% black, 46% white, and 2% other) from a university-based family medicine practice. Mean (SD) age was 51.9 years (8.9 years). Participants completed the Pittsburgh Sleep Quality Index, the Center for Epidemiologic Studies Depression Scale, Revised, and a checklist of chronic health conditions. RESULTS: The rate of clinically meaningful sleep disturbance was 71%. In bivariate logistic regressions, black race was associated with sleep disturbance (odds ratio [OR], 3.00; 95% CI, 1.17-7.69). Controlling for income attenuated that association by about 11% (race OR, 2.71; 95% CI, 1.04-7.06). Education explained about 35% (race OR, 2.39; 95% CI, 0.89-6.42). Adjustment for depression, chronic illness, and education simultaneously resulted in an estimate for race of OR, 2.44; 95% CI, 0.85-7.01. CONCLUSION: Being black is associated with a sleep disturbance that is accounted for only partially by depression, socioeconomic status, and disease burden. Black primary care patients may benefit from additional screening and monitoring of sleep difficulties.